TIOSPIR® (NCT01126437) was a large (N=17135), long-term (2–3-year), randomized, double-blind, parallel-group, double-dummy, event-driven trial in patients with COPD comparing the safety and efficacy of once-daily tiotropium (SPIRIVA® [Boehringer Ingelheim, Ingelheim am Rhein, Germany]) via Respimat® 5μg (2 inhalations of 2.5μg), and 2.5μg (2 inhalations of 1.25μg), or via HandiHaler® 18μg; 17116 patients constituted the as-treated population and were included in this analysis. The study design, detailed methods, and primary results were published previously.12,14

All participants had a confirmed diagnosis of COPD (forced expiratory volume in 1 second [FEV1] ≤70% predicted and FEV1/forced vital capacity ≤0.70), were aged ≥40 years, and had ≥10 pack-years of smoking history.14 Patients with concomitant cardiac diseases were included in the study, except those with recent (≤6 months) myocardial infarction (MI), severe arrhythmia, or a change in drug therapy within the previous year; or hospitalization for cardiac failure within the previous year.14 All usual background treatments for COPD, except other inhaled anticholinergics, were allowed.14

The geographical region of Latin America enrolled patients in Argentina, Brazil, Colombia, Guatemala, Mexico, Panama, and Peru.

Outcomes of interest in this analysis were time to all-cause death, time to first moderate-to-severe or severe exacerbation, time to major adverse cardiovascular events (MACE) and fatal MACE, and SAEs for patients in Latin America versus RoW.

An independent mortality adjudication committee, blinded to treatment assignments attributed the cause of each death.14 Non-fatal stroke and MI (included in MACE) were reported by study investigators and verified for classification accuracy by central reviewers who were blinded to treatment assignment.

Exacerbations were defined as the worsening of ≥2 major respiratory symptoms (dyspnoea, cough, sputum, chest tightness, or wheezing) for ≥3 days and requiring specified treatment changes. Moderate-to-severe exacerbations required a prescription for antibiotics, systemic corticosteroids, or both (with no hospitalization); severe exacerbations required hospitalization. The onset of exacerbation was defined as the onset of the first recorded symptom; the end of exacerbation was decided by the investigator, based on clinical judgment.

A composite endpoint of MACE was included in the analyses, comprising stroke, transient ischaemic attack (TIA), MI, sudden death, cardiac death, sudden cardiac death, or fatal event in the system organ classes (SOCs) for cardiac and vascular disorders. SOCs were defined according to the Medical Dictionary for Regulatory Activities standardizing attributions of AEs globally.

Based on similar efficacy and safety results among the tiotropium treatment arms in the primary analysis,12 data were pooled for this post-hoc exploratory analysis.

Baseline characteristics discriminating between patients in Latin America and RoW were identified and compared descriptively.

Hazard ratios (HRs) and 95% confidence intervals (CIs) for time to events by region were calculated using a Cox proportional hazards regression model without covariate adjustment, and displayed via Kaplan–Meier plots. A negative binomial regression model was used to estimate the number of exacerbations. HRs for time to death and time to moderate-to-severe exacerbation analyses, by region and baseline characteristics, are shown as forest plots. Additional analyses of time to death and moderate-to-severe exacerbation, severe exacerbations, and MACE were adjusted for covariates including age, post-bronchodilator FEV1% predicted, gender, exacerbation history, smoking history (pack-years), body mass index (BMI), and history of cardiac disorders, MI, coronary artery disease/ischaemic heart disease (CAD/IHD), cardiac arrhythmia, heart failure (class I–IV), and stroke/TIA.

Incidence rates, incidence rate ratios (IRRs), and 95% CIs for causes of death were calculated. SAEs are provided in the Supplementary Material.

For the mortality analysis (vital status), data were included if death occurred up until the study end date, irrespective of treatment discontinuation. For the exacerbation analysis (on-treatment), events were counted from randomization to drug stop date+1 day, and for MACE to drug stop date+30 days.

Of the TIOSPIR® population, 1000 patients from Latin America and 16116 patients from RoW were included (Table 1). Total exposure to tiotropium was 1855 and 32229 patient-years for Latin America and RoW, respectively. A total of 228 (22.8%) patients from Latin America and 3689 (22.9%) patients from RoW discontinued prematurely from trial medication. Reasons for premature discontinuation were similar, and mainly due to AEs, with respiratory disorders (including COPD) being the most common AEs, followed by neoplasms, and infections and infestations.12 Vital status follow-up was complete for 99.1% and 99.8% of patients from Latin America and RoW, respectively.

At baseline, patients enrolled in Latin America were comparatively older (higher proportion of patients aged ≥70 years), and had higher mean pack-years of smoking history than those in RoW, although a lower percentage were current smokers (Table 1). A higher proportion of patients in Latin America had COPD exacerbation in the year prior to the trial than those in RoW. Latin American patients were more likely to have very severe COPD (Global Initiative for Chronic Obstructive Lung Disease [GOLD] Stage IV) versus RoW (Table 1). Fewer Latin American patients had a history of cardiac disorders, or were receiving cardiovascular medication at baseline versus RoW (Table 1). The proportions of patients receiving long-acting muscarinic antagonist (LAMA) or long-acting β2-agonist (LABA) at baseline were lower in Latin America than in RoW; however, they were more likely to receive inhaled corticosteroids (ICS) (with or without LABA) (Table 1).

During the study, a greater proportion of patients died in the Latin American population (n=100 [10.0%]) versus RoW (n=1200 [7.4%]) (Table 2A). Analysis of time to death confirmed that patients in Latin America had a higher risk of death than those in RoW (HR [95% CI]: 1.52 [1.24–1.86]) (Table 2A; Fig. 1). This was also true when the analysis of time to death was adjusted for age, gender, FEV1% predicted, exacerbation history, smoking history (pack-years), BMI, history of cardiac disorders, MI, CAD/IHD, cardiac arrhythmia, heart failure (class I–IV), and stroke/TIA (HR [95% CI]: 1.37 [1.11–1.69]) (Table 2A).